Colorectal cancer is one of the leading causes of cancer mortality in the United States today. Although the underlying etiology is still not fully understood, it is increasingly apparent that genetic susceptibility plays a prominent role in a subset of these patients. Hereditary non-polyposis colon cancer (HNPCC) may account for the majority of the inherited forms of colon cancer, and it is estimated that as many as 1 out of every 200 individuals in the general population may be a carrier. This frequency makes HNPCC one of the most common inherited genetic disorders in humans. For a variety of reasons, including small family size and unavailable medical information, this diagnosis can be difficult to establish. Therefore, a reliable method is needed to define individuals who are at greatest risk for development of colon cancer and who would benefit most from aggressive management. Until recently, a strong family history of cancer was the only measure available to the clinician to decide which patients were at great risk. However, two types of molecular genetic markers have recently been identified that may enhance our ability to define colon cancer risk for families and individuals. These include the demonstration of tumor microsatellite instability ("mutator phenotype"), which is indicative of defective DNA mismatch repair, and the direct analysis of the genetic susceptibility loci responsible for this particular phenotype, hMSH2, hMLH1, hPMS1, and hPMS2. Importantly, as a result of these discoveries, the opportunity now exists to better define the natural history of this disease and to develop more useful clinical and molecular tools for the identification and pre-symptomatic diagnosis of those individuals at high risk for developing cancer. Because of the multitude of mutations that can occur within these genes, additional data on implications for phenotype are urgently needed. This data must be collected on population-based series of cancer patients to avoid the potential bias in risk that can arise from the study of HNPCC families ascertained because of their exceptional family history of cancer. Given a pressing need to address these issues, we have focused our attention on methods that would help identify those individuals who are most likely to benefit from genetic testing for heritable cancer risk, and on the characterization of both somatic and germline alterations within the genetic susceptibility loci. Our specific aims, therefore, are to: 1) assess the association of family history with tumor microsatellite instability; 2) estimate the frequency of germline mutations and define the spectrum of mutations in the four genetic susceptibility loci among patients demonstrating tumor microsatellite instability and/or a positive family history; 3) develop and evaluate a questionnaire that is effective in identifying patients who are most likely to have a genetic predisposition for cancer; and 4) determine the frequency of tumor microsatellite instability in an Alaska Native population.